PrismML Releases Bonsai 2: Efficient Open Models for Local and Edge AI Agents
A post says PrismML released Bonsai 2 with 2B, 4B, and 9B parameter models based on Qwen3.5 using ternary quantization.
TLDR
A post says PrismML released Bonsai 2 in 2B, 4B, and 9B sizes under Apache 2.0, with model files of 430MB, 950MB, and 2GB, respectively. The models are designed to improve reasoning, visual understanding, instruction following, and tool calling. According to the post, PrismML reports that the 9B model uses roughly 40% fewer tokens on standard instruction-following and everyday math tasks; file size alone does not establish the memory needed to run it.
Combined views
—
2 Sources, first seen ago